Pretreatment Method of Deep Learning
Method for Target Detection in Sea Clutter
Yanan Xu, Ningbo Liu
(&) , Hao Ding, Yonghua Xue, and Jian Guan
Naval Aviation University, Erma Road 188, Yantai 264001, Shandong, China
{yt_yananxu,lnb198300}@163.com
Abstract. Sea target detection is widely used in military and civilian fields.
Because of the space-time non-stationary characteristics of high resolution radar
sea clutter, traditional target detection methods have many limitations and are
limited by the use of scenarios. In recent years, with the progress of deep
learning in image classification tasks, a series of target detection methods based
on deep learning have emerged. By applying these methods to target detection in
radar sea clutter, high accuracy and good generalization can be obtained.
However, there are many new problems when these target detection methods,
mostly based on computer vision, are introduced to target detection in radar sea
clutter due to different data forms and detection standards. This paper mainly
discusses the effects of different pretreatment modes of target detection in sea
clutter using classification target detection framework on training speed and
detection accuracy.
Keywords: Deep learning Á Target detection Á Sea clutter Á False alarm rate
1 Introduction
Effective detection of sea targets under sea clutter background is of great significance in
both military and civilian fields. In order to meet the needs of target detection, radar
needs to have high resolution in range and azimuth; high resolution sea clutter has nonstationary characteristics, and contains a large number of sea peaks, which seriously
affects the accuracy of target detection. At the same time, traditional adaptive matched
filtering methods need clutter to meet certain model assumptions and have a small
application range [1]. Although the method based on fractal theory is simple and
efficient, it needs to accurately estimate the interval with fractal characteristics and
needs long observation time [2]. In the actual detection of sea targets, only for the
specific target type under the background of specific clutter has a good detection
accuracy. Since AlexNet [3] put forward in 2012, a large number of target detection
algorithms based on deep learning have been produced, including classification-based
detection algorithm based on convolutional neural network (CNN) and regressionbased target detection framework [4].
Deep learning target detection methods are mostly proposed for object detection in
two-dimensional images in computer field, which are often used in face recognition,
automatic driving, etc. [4]. Although radar display screen has different visualization
forms such as A-display and P-display, they are all obtained by radar echo signal
© Springer Nature Singapore Pte Ltd. 2020
Q. Liang et al. (Eds.): Artificial Intelligence in China, LNEE 572, pp. 608–614, 2020.
https://doi.org/10.1007/978-981-15-0187-6_72
Method for Target Detection in Sea Clutter
Yanan Xu, Ningbo Liu
(&) , Hao Ding, Yonghua Xue, and Jian Guan
Naval Aviation University, Erma Road 188, Yantai 264001, Shandong, China
{yt_yananxu,lnb198300}@163.com
Abstract. Sea target detection is widely used in military and civilian fields.
Because of the space-time non-stationary characteristics of high resolution radar
sea clutter, traditional target detection methods have many limitations and are
limited by the use of scenarios. In recent years, with the progress of deep
learning in image classification tasks, a series of target detection methods based
on deep learning have emerged. By applying these methods to target detection in
radar sea clutter, high accuracy and good generalization can be obtained.
However, there are many new problems when these target detection methods,
mostly based on computer vision, are introduced to target detection in radar sea
clutter due to different data forms and detection standards. This paper mainly
discusses the effects of different pretreatment modes of target detection in sea
clutter using classification target detection framework on training speed and
detection accuracy.
Keywords: Deep learning Á Target detection Á Sea clutter Á False alarm rate
1 Introduction
Effective detection of sea targets under sea clutter background is of great significance in
both military and civilian fields. In order to meet the needs of target detection, radar
needs to have high resolution in range and azimuth; high resolution sea clutter has nonstationary characteristics, and contains a large number of sea peaks, which seriously
affects the accuracy of target detection. At the same time, traditional adaptive matched
filtering methods need clutter to meet certain model assumptions and have a small
application range [1]. Although the method based on fractal theory is simple and
efficient, it needs to accurately estimate the interval with fractal characteristics and
needs long observation time [2]. In the actual detection of sea targets, only for the
specific target type under the background of specific clutter has a good detection
accuracy. Since AlexNet [3] put forward in 2012, a large number of target detection
algorithms based on deep learning have been produced, including classification-based
detection algorithm based on convolutional neural network (CNN) and regressionbased target detection framework [4].
Deep learning target detection methods are mostly proposed for object detection in
two-dimensional images in computer field, which are often used in face recognition,
automatic driving, etc. [4]. Although radar display screen has different visualization
forms such as A-display and P-display, they are all obtained by radar echo signal
© Springer Nature Singapore Pte Ltd. 2020
Q. Liang et al. (Eds.): Artificial Intelligence in China, LNEE 572, pp. 608–614, 2020.
https://doi.org/10.1007/978-981-15-0187-6_72